The Third Test: Statistically Significant Other

For all that the prior two tests were successful, they didn’t tell me very much. To start with, all of this could have been faked by Father and his associates. Lizzie, cute as she is on the pillow next to mine, might very well have been paid to sleep with me. Just the thought of it hurts my pride, but I doubt that it’s true—from the way Dave put it, Lizzie was already at the bar before I even left the house, and they couldn’t have known where I was going.

I check the app anyways, though:

Elizabeth Kestrel, Score: 90

All seems to be in order. She’s only fallen harder for me since we left the bar.

The bigger problem is that the tests might have just been passed through sheer coincidence. Three out of four chocolates on the bagel trick might have been better than average, but it was only by the smallest possible margin. Lizzie sleeping with me was just par for course. No, the real reason why I did these two tests was to pass the time while I waited on the results from my third.

I pull out my phone and open up Tinder. This is the real test.

Early this morning, I downloaded my Tinder data and had a sobering look at my track record in the online dating sphere.

Over the past couple years, I’ve swiped right on precisely 1,037 profiles. Out of those, 260 swiped back on me—roughly a quarter of the total. If that proportion seems high to you, it’s because I’m fairly picky with who I swipe on in the first place. I’ve developed a pretty good sense of who might be receptive to my profile, and so a decent number of girls bite. I’m proud of my ratio.

If that proportion seems low to you, fuck you, and let me know what you’ve been doing differently.

Early this morning, I also went through Tinder and started swiping like a madman. I can’t tell you how many dozens of profiles I went through, cute faces and mirror selfies blending into one another until it was just a morass black and white and Latino and Asian faces that I was discarding on autopilot, but after I’d seen two joke profiles featuring the Statue of Liberty, I knew I’d gone far enough. Everyone coming up now would be a clean potential match. If I swiped on them, then they only see my profile after the Affection Multiplier had taken effect.

See, if the Affection Multiplier was really working, I would expect to have a match rate of 50% now—double that of the quarter I used to have. Now, I’m going to throw some large sample sizes and statistics at this thing and figure out whether it’s the real deal.

Early this morning, I swiped right on 30 profiles. Now, I look at the app, and see that I’ve been matched with 13. That’s 43%. My heart leaps in my throat, knowing that this is a good sign, but I’m not ready to celebrate just yet.

I pull out Excel and run the math. Fisher’s Exact Test gives me the probability that this difference in proportions that I’m seeing is statistically significant—that the difference between 43% and 25% that I’ve observed is real, and not just due to random statistical noise. I run the numbers. I need the p-value to come out beneath 5%—for the chance of a false positive to be lower than that.

The answer prints out on the screen.

p = .024

That’s a 2.4% chance. I’m in the clear.

“Fuck yes,” I cry out, throwing my hands in the air.

The Affection Multiplier is real. The shit works. I dare to let myself dream for a second, imagining myself kicking back in a hot tub in the Alps, caviar on the side, my beautiful sister Kara lounging in the tub next to me, enjoying the view out of the large bay windows and the snow-covered peaks that lay beyond…

There’s a rustling at my bedside. My eyes snap down to Lizzie, who I realise has been stirred awake by my exclamation. I watch as her eyes take in my glowing phone, still left on the bed, still showing my Tinder matches, and then a primal snarl crosses her face.